Dynamic programming (DP) models have been proposed by a number of researchers to develop optimal reservoir operation. In this paper, we propose to eliminate the fixed threshold values, specifically the maximum allowable release and reservoir storage from the DP model and replace them with dynamic dependent values calculated by a fuzzy inference engine. For this purpose, a simple model of DP is modified based on a novel intelligent state dropping (ISD) mechanism. The ISD mechanism is designed based on fuzzy logic theory. Although the proposed methodology is widely used for broadband satellite based internet protocol network congestion control, its application in water resources management has not been reported to date. Application of the ISD mechanism incorporates the inflow uncertainty in reservoir optimization model in a simpler way than the stochastic DP-based optimization models (SDP). Furthermore, the multi-purpose objective function of DP-model is changed to a single objective function. The simulation results showed that the newly proposed model reduces the shortfall in supplying demand and improves the reservoir operation performance indices, i.e., the reservoir reliability indices, as compared with the results by the DP and SDP models.
Keywords: fuzzy logic, optimal reservoir operation, state dropping mechanism, uncertainty